Frequently, roads and railways are co‐aligned along the same corridor (e.g., Proctor
et al. 2005; Li et al. 2010). In fact, the co-occurrence of roads and railways is an
important aspect to have in consideration, as wildlife response to one infrastructure can
condition the response to the other. For example, in a study in the USA, Waller and
Servheen (2005), found that radio-collared grizzly bears (Ursus arctos) crossed a
highway-railway corridor at night, presumably to avoid high diurnal highway traffic,
but when railway traffic was heavier, there was a behavior that led to higher mortality
rates on the railway than on the road. In another study in Canada, Clevenger and Waltho
(2005) found that black bears (U. americanus) and cougars (Puma concolor) tended to
cross a highway along wildlife passes far from railway tracks, while wolves (Canis
lupus) preferred to use crossing structures close to the railway.
It is important to acknowledge the different ways that roads and railways impact
wildlife. We consider the following five to be the most relevant: First (1) Traffic
flow is much lower on railways, with several case studies showing that the number
of trains moving through railways per unit of time are about 0.2–1.6% of the
number of cars moving through nearby roads (Gerlach and Musolf 2000; Waller
and Servheen 2005; Xia et al. 2007; Kušta et al. 2015); (2) railway traffic flow is
characterized by long traffic-free intervals—in some cases there is no nighttime
railway traffic (e.g., Rodríguez et al. 1997; Pérez-Espona et al. 2008)—although
noise and vibrations produced by trains are higher than those produced by cars
(Dorsey et al. 2015; Kociolek et al. 2011); (3) railways have lower wildlife mortality, possibly because of lower traffic flow (Cserkész and Farkas 2015; Kušta et al.
2015), but this assertion should be taken carefully, because figures for roads may be
inflated as they are far more widespread than railways (Pérez-Espona et al. 2008;
Yang et al. 2011) (4) railway corridors are narrower than those of roads, which
imply lower loss of habitats when a new line is built (e.g., Gerlach and Musolf
2000; Tremblay and St. Clair 2009); finally, although some maintenance practices
that use pollutants (potential disturbers) are shared by railways and roads, like
de-icing or the application of herbicides on verges, the impact of vehicles, the most
important source of chemical pollutants (Forman et al. 2003), is lower in railways
because many trains have electric engines.
Especially interesting for our purposes are the studies comparing railway impacts
with those of roads, as they highlight their similarities and differences. These differences and similarities, however, tend to be species-specific. For instance, Gerlach
and Musolf (2000) found that in Germany and Switzerland, a 25-year-old highway
contributed to genetic substructuring in bank voles (Clethrionomys glareolus), while
a 40-year-old railway and a 25-year-old country road did not. In a similar vein,
railways seemed to have no strong effects on the red deer (Cervus elaphus) population’s genetic differentiation in the UK, as differentiation was the same with or
without railways, while roads were identified as gene-flow barriers to this species
(Pérez-Espona et al. 2008). Railways had a low impact is the latter example probably
because they were not parallel to orographic barriers, they were relatively sparse, and
they had a low traffic flow (Pérez-Espona et al. 2008). In tune with the previous
work, Yang et al. (2011) found that roads contributed to the genetic isolation of
Chinese populations of Przewalski’s gazelles (Procapra przewalskii), but railways
6
L. Borda-de-Água et al.
et al. 2005; Li et al. 2010). In fact, the co-occurrence of roads and railways is an
important aspect to have in consideration, as wildlife response to one infrastructure can
condition the response to the other. For example, in a study in the USA, Waller and
Servheen (2005), found that radio-collared grizzly bears (Ursus arctos) crossed a
highway-railway corridor at night, presumably to avoid high diurnal highway traffic,
but when railway traffic was heavier, there was a behavior that led to higher mortality
rates on the railway than on the road. In another study in Canada, Clevenger and Waltho
(2005) found that black bears (U. americanus) and cougars (Puma concolor) tended to
cross a highway along wildlife passes far from railway tracks, while wolves (Canis
lupus) preferred to use crossing structures close to the railway.
It is important to acknowledge the different ways that roads and railways impact
wildlife. We consider the following five to be the most relevant: First (1) Traffic
flow is much lower on railways, with several case studies showing that the number
of trains moving through railways per unit of time are about 0.2–1.6% of the
number of cars moving through nearby roads (Gerlach and Musolf 2000; Waller
and Servheen 2005; Xia et al. 2007; Kušta et al. 2015); (2) railway traffic flow is
characterized by long traffic-free intervals—in some cases there is no nighttime
railway traffic (e.g., Rodríguez et al. 1997; Pérez-Espona et al. 2008)—although
noise and vibrations produced by trains are higher than those produced by cars
(Dorsey et al. 2015; Kociolek et al. 2011); (3) railways have lower wildlife mortality, possibly because of lower traffic flow (Cserkész and Farkas 2015; Kušta et al.
2015), but this assertion should be taken carefully, because figures for roads may be
inflated as they are far more widespread than railways (Pérez-Espona et al. 2008;
Yang et al. 2011) (4) railway corridors are narrower than those of roads, which
imply lower loss of habitats when a new line is built (e.g., Gerlach and Musolf
2000; Tremblay and St. Clair 2009); finally, although some maintenance practices
that use pollutants (potential disturbers) are shared by railways and roads, like
de-icing or the application of herbicides on verges, the impact of vehicles, the most
important source of chemical pollutants (Forman et al. 2003), is lower in railways
because many trains have electric engines.
Especially interesting for our purposes are the studies comparing railway impacts
with those of roads, as they highlight their similarities and differences. These differences and similarities, however, tend to be species-specific. For instance, Gerlach
and Musolf (2000) found that in Germany and Switzerland, a 25-year-old highway
contributed to genetic substructuring in bank voles (Clethrionomys glareolus), while
a 40-year-old railway and a 25-year-old country road did not. In a similar vein,
railways seemed to have no strong effects on the red deer (Cervus elaphus) population’s genetic differentiation in the UK, as differentiation was the same with or
without railways, while roads were identified as gene-flow barriers to this species
(Pérez-Espona et al. 2008). Railways had a low impact is the latter example probably
because they were not parallel to orographic barriers, they were relatively sparse, and
they had a low traffic flow (Pérez-Espona et al. 2008). In tune with the previous
work, Yang et al. (2011) found that roads contributed to the genetic isolation of
Chinese populations of Przewalski’s gazelles (Procapra przewalskii), but railways
6
L. Borda-de-Água et al.